Again, the process of earning a living has gotten in the way of the process of having fun, and I haven't done much to the turbine for a couple of months or so. However, I recently spend a couple of evenings building a prototype of a liquid fuelled burner system for the engine, and also stripped, cleaned, painted, and rebuilt the turbine itself. I also replaced the oil pump, rebuilt the oil tank and feed system, built an integrated starting system for the engine, and changed the oil. It was a busy couple of evenings.
It has been very dull and wet weather lately, and as a result the pictures on this page are a bit washed out, due to bad lighting. However, they are still very viewable, just not bright and sparkly and cheerful. Sorry.
Here are three different views of the cleaned and repainted engine, and a shot of the various sub-assemblies that go to make up the turbocharger. I cleaned and painted the turbine housing with car exhaust enamel, mainly to protect it from rust. The red colour was picked because it was the only one the shop had in stock. Since I had painted the turbine housing, I naturally had to paint the compressor housing and CHRA as well, or it would have looked silly. The overall colour scheme looks reasonably good, but I don't expect it will last long once I start running the engine. The exhaust enamel is supposed to be able to stand a temperature of 650 deg. C continuously, but it went a funny colour after about 5 minutes running from the rebuilt engine, and flaked a bit when it cooled down. Oh well.
As you can see, I have temporarily removed the afterburner vapouriser, as it got in the way when I was painting the turbine housing. I am going to make a new injector nozzle out of 3mm stainless steel tube, and put it back on at a later date. It will be a simpler shape than the original circular nozzle, and will inject the fuel closer to the centre-line of the exhaust nozzle, which may give slightly better results.
You can see the new oil pump, which is one from a Mk II Ford Escort, with an adaptor plate made from 6mm mild steel which allows 1/4 BSP fittings to be used. I got the pump for 3 UKP from a local scrapyard, along with two 12V radiator fan motors for 1 UKP each. I have used one motor for driving the oil pump, and the other one for driving the fuel pump. The oil pump takes quite a lot of power, and the motor pulls about 8 amps when the oil is cold. The current requirement drops quite a bit once the oil has warmed up, though. I have started using a 0W40 high performance racing oil, which will probably give better lubrication than the home-made mix I was using before, but is quite a bit more viscous. When it's cold, the turbine spins up very sluggishly, and I have found that I need to preheat the oil and run the turbine briefly without the pump switched on to get it to start properly. This situation may improve once the weather warms up again. Below is a close-up of the finished pump, complete with filter. It can produce a reasonable flow at up to about 6 bar or so.
I have now completed the liquid fuelled burner for the combustor, and it sort of works. The burner is based around standard oil-fired furnace parts, obtained from a local heating supplier, who has been very helpful. In fact, it turned out that the owner of the company, and his son, are both very interested in turbine engines, and I have had a lot of useful advice from them on oil burners.
The nozzle I am currently using is a 1.5 gph Danfoss one that cost 7.50 UKP. It is mounted to the base plate of the combustion chamber via a brass adaptor that terminates in a 1/4 BSP fitting. This has a 3/8 BSP female hydraulic quick release coupling connected to it on the other side of the base plate, via a 1/4-3/8 BSP adaptor. There is also an auxiliary propane nozzle fitted, which is basically the same as the one in my original design. It is arranged so that the flame from it passes over the oil nozzle at an angle, at a distance of about 2 cm. This is used for starting and preheating the engine, and as a boost fuel system. It also terminates in a quick release fitting, this one a 1/4 BSP PCL female coupling.
The new flame holder is considerably shorter that my original one, and rather larger in diameter. It is designed to allow more air to enter at the base and mix with the liquid fuel early on, to hopefully produce a shorter, hotter flame, and more complete combustion before the turbine inlet. Below are two images of the burner without the flameholder fitter, one from each side, and one of the fully assembled system.
To go with the above modifications, I built a starting and control system for portable operation. It's built into a case that originally housed an oxygen resuscitation system from an ambulance. This box has a lot of stainless steel tubing and fittings, all 1/4 BSP and compatible with the ones I have been using throughout the turbine project. Three cheers for standards! Anyway, I removed the original system, and built in the fuel pump and motor, fuel tank, battery, a small propane cylinder, and the control system, reusing most of the oxygen fittings. Actually, I haven't yet fitted the fuel tank and control panel yet, as I haven't finished making the panels. Here are a couple of pictures of the current version of the control box.
The propane bottle and valve are from a small propane torch that I sacrificed to the cause. Again, it had a 1/4 BSP thread on the valve, once I removed the burner. This mounted neatly through the hole in the box where one of the original oxygen fittings was. The high pressure hydraulic female coupling went through the other hole, and just has clearance from the fitting on the propane system. A couple of 2m hoses with the appropriate fittings ling the control system to the engine. The oil pump can just be seen under the U shaped loop of tubing at the top of the box in the first picture. It is again made by Danfoss, and cost 15 UKP second-hand. I was originally going to use another, larger pump I already had, but it turned out to be too big to fit into this box. In the interests of neatness, I decided to get a smaller pump. The new pump is about 5cm on a side, can produce pressures of about 20 bar, and has a flow rate of up to 10 gph or so, according to the heating people. It takes a surprisingly small amount of power from the motor, which is coupled to it via a short piece of thick walled PVC tubing, allowing for shaft misalignment. The current consumption of the motor at full speed on 12V is only about half an amp, and varying the motor voltage changes the pressure quite well. The pump also has a fuel cutoff valve, operated by a solenoid with a removable coil. The original coil was 240V AC, and I had to replace it with a 12V one.
The somewhat complicated plumbing around the pump consists of a pressure dump valve (nearest the side of the box), which connects the input and output of the pump together when open, a needle valve between the pump output and the outside world, and the fuel suction hose, which goes to the tank. The dump valve is needed to relieve the pressure in the output fuel hose when the pump is stopped, otherwise it dribbles incontinently for a while. It can also be used for coarse throttling of the fuel flow. The needle valve was intended to be a fine control throttle, but doesn't seem to work very well. It may in fact be faulty, and I will get around to checking it at some point in the near future.
The battery is a sealed 24 A/hr 12V lead acid one, mounted just behind the fuel system, to balance the box. The fuel tank will be mounted behind this, and will hold about 2.5 l of fuel. The propane bottle holds about 0.5l of gas, and will run the turbine at a fast idle (about 20000 rpm) for several minutes. The main problem with it is that it frosts up very rapidly, and the vapour pressure falls off. The control panel will consist of switches for activating the oil and fuel pump motors, the ignition, the pump solenoid, and eventually the air starter and propane valves. It will also have a tachometer display, and temperature gauges for the EGT and TIT, as well as oil and fuel pressure gauges. Much of this is still a little way off. Ultimately, I will have a system that allows complete remote operation of the turbine, and may well incorporate an engine management system allowing push-button starting and running.
Larger versions of the above pictures can, as usual, be downloaded below.
The results of the initial tests of the completed system were mixed. On the one hand, it didn't blow up, quite. On the other hand, it didn't work properly either. Starting the engine on propane with the new burner was the first task. This proved to be a problem at first, because I made the mistake of turning on the oil feed first. The oil viscosity was sufficiently high when cold to cause so much drag that the turbine wouldn't spin up. Flushing the bearing through with WD40 cured this, and the thing then ran up quite well, although not as smoothly as with the original burner. This is probably due to the new flameholder not having the optimal hole pattern/density, which will need some tweaking. Once it had warmed up for a few seconds, the oil feed was turned on and proved to mostly work. It also proved that the bearings on the current turbocharger leak like sieves, which will have to be rectified. The turbine shaft looked OK when I stripped the engine, and I am going to try to get a new set of bearings for it, which will hopefully work better.
Once the turbine was idling smoothly, I cautiously turned on the kerosine feed at midrange flow. The turbine instantly accelerated at an incredible rate, doubling its speed within a second or two, accompanied by a loud farting roar and a metre long orange flame out of the exhaust. This nearly gave me heart failure, and I immediately turned it off, in the process of falling backwards in shock. I now have a graphic idea of why an overfuel situation is called turbine runaway, because that's what you scream when it happens. "Run away, run away".
After having a little lie down to recover, I tried again. This time I tried using minimum flow, and throttling with the pressure dump valve. This produced a very disappointing result, merely giving a slight speed up, and immediate flame-out when the propane was turned off. A few more tests, and it transpired that the kerosine feed will dramatically boost the engine speed when it's over a certain threshold level, but is unstable in operation without the propane on as well. I'm not yet sure why. I did manage to get it to self-sustain on kerosine alone once or twice, but the thing either just kept accelerating or flamed out. The sound effects were amazing. During acceleration the exhaust flame is enormous, and crackles violently. If the fuel flow is cut a bit, it flames out, smokes like mad, and sometimes reignites briefly with an upsettingly loud boom. This is completely different behaviour than my original combustor produced, even at comparable turbine speeds. I gave up after about an hour or so, mainly because I was getting worried that I would either overheat the turbine or produce a detonation so loud the police would turn up.
I obviously have an air flow problem of some sort. The next stage is to remove the combustor and fire it up with an external air source, so I can see why the combustion is so unstable. I suspect that my flameholder needs quite a bit of work, and is producing either too weak or too rich a mixture, or possibly both depending on the engine speed. I may have been lucky in my choice of the number, position, and size of the holes in the first one I made. I also need to replace either the bearings or the complete turbocharger, as I am getting tired of large amount of hot oil spraying all over the place.
Back to the beginning of turbines. Back
